January 21, 2004

Le Grand's Gambit

Julian Le Grand is Professor of Social Policy at the London School of Economics. His latest book, Motivation, Agency, and Public Policy: Of Knights & Knaves, Pawns & Queens, is an examination of the implicit and explicit assumptions about human behavior that underly the public policy prescriptions of both ends of the political spectrum.

As the title suggests, Le Grand posits two axes upon which to chart these assumptions: motivation and agency. He then introduces the "characters" of the subtitle to label the extremes of the axes. "Knights" are motivated by altruism and the spirit of cooperation, while "knaves" are self-interested utility maximizers — the kind of actors touted by public choice theorists. Agents are further divided into passive "pawns", cogs in the public service machine, and active "queens", empowered to make choices and free to act.

He demonstrates that this framework is interesting by examining how various policies promoted by the Left and Right, or as he calls them, the social democrats and neo-liberals, make different assumptions about where people fall (or should fall) on these axes. Social democratic policies tend to assume that service providers (politicians, bureaucrats, public servants) will be knightly while benficiaries will be pawns. Neo-liberals assume the opposite: producers are knaves and consumers should be queens.

Unfortunately, he continues in this vein and treats the two axes fairly differently, almost as two separate, one-dimensional explanations, focusing on motivation for service providers and agency only for recipients. It would be interesting to have the framework fleshed out to examine how both roles can move in this two-dimensional space.

In addition, while he takes a descriptive approach to the motivation axis, his treatment is almost completely normative with respect to agency. He takes it as given that policies should empower citizens and treat them as active agents in their implementation. While libertarians and classical liberals would certainly agree with this assertion, it seems to beg the question to some extent — those who believe that the collective good must come before individual choice may reject this starting point.

Le Grand could have bolstered his argument by showing that policies must take into account individual agency to be effective, i.e. a descriptive argument that showed how policies that treat citizens as pawns are likely to fail. The normative argument could have followed strongly on this consequentialist base.

This conceptual framework is laid out in the introduction. The three subsequent sections focus on Motivation, Agency, and Policies, respectively.


As mentioned above, his theory of motivation focuses mainly on the service providers, examining how they fall into the categories of altruistic "knights" and self-interested "knaves".

He describes how communal spirit after the Allied victory in WWII led, at least in the UK, to the renewed belief in man's ability to set aside selfishness and act in the interest of the community. This, in turn, led to the great movement in social democracy in England in the 50's and 60's. Studies and literature from the period claimed that not only were systems that relied on voluntary cooperation and altruistic motives more desirable from a normative point of view, they were actually more efficient. The most famous of these is Richard Titmuss' study of blood donors in Britain's all-volunteer system and his comparison to "market" systems as used in America.

Failures of these systems to live up to the promises (including the eventual need to import large quantities of blood from America) led in the late 70's and 80's to their replacement by quasi-market systems. These programs, brough about as part of the Iron Lady's conservative reforms, attempt to appeal to providers' self-interest to increase the quality, quantity, and efficiency of public services provided.

Le Grand then goes to the evidence to determine which kinds of programs actually are more effective. He starts from the assumption that, as long as they work, programs that promote and rely on altruism should be considered morally superior and better for society. He brings to bear some evidence from studies that show that market forces can actually remove people's willingness to contribute to the greater good — the collectivist fear that markets corrupt mankind and are, in fact, the source of, rather than the solution to, selfishness. He contrasts this with the hard-to-deny efficiency of market systems, particular in comparison to the failing social democratic policies of the 70's. He also cites studies that show that some incentives can increase altruistic behavior in some circumstances, including one of home care professionals in Britain who were given a small stipend for their time.

From these interesting but conflicting studies, he comes up with a compelling, but unfortunately very speculative, argument for an S-shaped supply curve for knightly service providers. He argues that small incentives will increase the supply of communal service as the implied recognition encourages knights to provide more services. At a critical threshold, he argues, the incentive overpowers the sense of sacrifice that drives the altruistic behavior and the supply actually decreases. Finally, full market-based incentives lead to an upward sloping supply curve based only on self-interested utility maximization. This is an attractive story, one I find plausible and compelling, but it seems to be based only on a small number of inconclusive and contradictory studies in different areas of "communal service". It would seem to be a promising area for future research, but it's also a tenuous foundation on which to build your theory of motivation.

Despite that, Le Grand tells an interesting story based on this theory and discusses some of the issues that arise because of the multiple equilibria that this "backward-bending" supply curve generates.

He ends by arguing for what he calls "robust" policies that encourage knightly behavior when possible, but also work in situations where knaves dominate. He leaves this as a vague idea in the theory section, but makes it (a bit) more concrete in the Policies section.


Le Grand's theory of agency focuses mainly on the recipients of public services, clustering them into active "queens" and passive "pawns". He starts from the liberal position that individual empowerment is a good, but he acknowledges the collectivist concern that it is not necessarily an unqualified good.

He offers an enlightening look at the problems that can arise from putting "too much" power in the hands of the individual. Here he combines some of the findings of behavioral economics with real world policy issues to show the kinds of failures that humans can be prone to.

Most of these failures are familiar arguments in favor of paternalistic policies of doing what's best for people and "protecting them from themselves". Le Grand, rather than always moving to limit choice because of these possible failures, hopes to find instituional ways to limit their effect. For example, he discusses the value of moving sources of knowledge closer to the agents to help mitigate "irrational" choices made due to limited information. And he encourages building in stuctural incentives to guard against moral hazard and adverse selection effects.

That said, his argument for forced savings, i.e. that the future selves of the individuals need to be respected and given a voice, while novel, seems more like a justification than anything truly explanatory.


The final section contains several concrete proposals that attempt to apply his theories of motivation and agency to specific public goods. While most of his proposals are crafted for the UK, his arguments are general enough to be of interest to anyone concerned with modern welfare state policies.

While the proposals are too extensive to cover in detail here, they are worth mentioning in broad brush strokes to show where the theories take him.

In education, he advocates a form of school choice, like vouchers, pointing to their success in the UK. But he adds in what he calls a "positive discrimination" voucher that would encourage schools to counteract the stratification that is the most compelling argument against vouchers.

In healthcare, he obviously start from the British perspective of a National Health Service, but again argues for increased choice in providers to help harness market-forces for the sake of efficiency. He has a really interesting discussion of the kinds of incentives that manifested themselves (both for doctors and patients) in the various systems put in place by Thatcher and subsequently changed by Labour. He also highlights some of the specific problems associated with empowering people with regard to healthcare — particular the knowledge problem (i.e. the fact that not everyone has a medical degree and can make informed choices as consumers). He discusses the role of the GP in providing information and facilitating good choices by setting up systems that ensure that GP's incentives are balanced between helping their patients first and foremost, and furthering the public good in aggregate.

In social security, he argues for matching funds from the government to foster a sense of partnership in providing for retirement. A graduated scale where the first dollars saved were matched one-to-one, while later dollars were matched at lower rates would ensure that the rich did not get the lion's share. Again, he struggles again with the knowledge problem of empowerment and questions how the savings should be invested and by whom, and under what circumstances they could be used early (e.g. to buy a house or start a business).

His final proposal is the most bold and controversal — the idea of demogrants, cash given to citizens either at birth or at majority. He would fund this from increases in estate taxes. While libertarians should scream bloody murder at the coerced redistribution of wealth, the trade-off is the increased liberty in determining how the money is spent by pushing the decision making down to the individual. The main problem that I see is that society is unlikely to give the money without having some say in how it is spent, and this promises to be an ever increasing set of restrictions (the flip side of tax breaks). Second, for this to achieve the desired results, squandering the demogrant would need to have real consequences and lead to some hardships, or again, the incentives work against you – it's just a freebee. But any nation that decides to redistribute the money to everyone is also unlikely to make people face the full consequences of misuse.

In all, a great book, though I wished for more in some parts. It's greatest contribution is similar to that of Postrel's The Future and Its Enemies in that it provides a new and useful way to examine policies and proposals. As I watched the State of the Union speech tonight, I found myself thinking about whether each policy would make citizens queens or pawns, whether each program assumed that bureaucrats were knaves or knights. I hope this framework sticks with me throughout the 2004 election.

Posted by richard at January 21, 2004 12:57 AM